Poster Board I-338
Platelet secretion defects are a common type of inherited platelet disorder. Experts agree that ATP release assays are often helpful to assess for platelet secretion disorders among individuals with suspected bleeding problems. However, the diagnostic utility of ATP release assays has not been established, and it is unclear if the findings correlate with aggregation results. Our goal was to evaluate the incidence and spectrum of platelet release abnormalities in a prospective cohort of individuals referred for testing as part of a bleeding disorder assessment, and describe the diagnostic utility of platelet ATP release and its relationship to aggregation findings.
Data were analyzed for 40 healthy controls and 76 patients from a prospective cohort of individuals referred for bleeding disorder assessments. ATP release and light transmission aggregometry (LTA) were assessed by standardized methods, with validated reference intervals (derived by non-parametric analysis), using platelet-rich plasma adjusted to 250 × 109 platelets L−1 with autologous platelet poor plasma. ATP release was assessed using the Chronolog lumiaggregometer, as recommended, and a modified agonist panel (1.6 mM arachidonic acid, 1 uM thromboxane analogue U46619, 100 uM epinephrine, 6 uM epinephrine, 5 ug·mL−1 collagen, 1.25·g mL−1 collagen, 1 U·mL−1 thrombin, and 5 uM ADP). The test sensitivity and specificity for inherited platelet disorders were estimated and the likelihoods of detecting an inherited platelet disorder by ATP release (using different agonists or the agonist panel) were expressed as odds ratios (OR) and 95% confidence intervals (95% CI). Correlations between ATP release triggered by different agonists, and between ATP release and maximal aggregation with the same agonist, were expressed as Pearson correlation coefficients (ρ).
Reduced ATP release with one or more agonists was more common among the referred individuals that a hematologist categorized as having a bleeding problem (n = 66; 45% abnormal) than among healthy controls (5% false positives, p < 0.0001). The likelihood of an inherited platelet disorder was high (OR 83, 95% CI 16 - 445) when ATP release was abnormal with one or more agonists, even when LTA results were normal (OR 56, 95% CI 10 - 318). Impaired ATP release with some agonists (e.g. 5 uM ADP and 100 uM epinephrine) could not be detected as the reference intervals did not have a measurable lower limit. Among other agonists, 1 U·mL−1 thrombin had the highest likelihood of detecting impaired ATP release due to an inherited platelet disorder (OR 32 compared to 17- 25 for other agonists).
Receiver operator curve analyses indicated that an assessment of ATP release with the full panel, or a more limited panel of 4 agonists, had high specificity and moderate sensitivity for detecting inherited defects in platelet function. The predictive power was improved by considering results as continuous rather than categorical (normal/abnormal) variables, however, this required complex mathematical equations. Most abnormalities in ATP release due to common inherited platelet disorders were detected by a limited number of agonists in the panel. The optimal panel, with minimized false positive and false negative rates (5% and 18% respectively), included: 1 uM thromboxane analogue U46619, 6 uM epinephrine, 5.0 ug·mL−1 collagen, and 1 U·mL−1 thrombin.
Among patient subjects and healthy controls, there was a significant correlation (ρ = 0.25 – 0.92; p values < 0.02) between the amount of ATP release induced by most agonists. Maximal aggregation and ATP release responses for each agonist did not correlate significantly among healthy controls, and showed only a weak correlation among patient subjects.
Like light transmission aggregometry, ATP release assays have important diagnostic utility for detecting impaired platelet function due to common inherited bleeding disorders. The poor correlation between ATP release and maximal platelet aggregation suggests that these tests are predominantly influenced by different parameters of platelet function. Our data on useful agonists for detecting impaired ATP release may prove helpful to laboratories that wish to optimize testing for common platelet disorders. The knowledge that platelet secretion assays are very helpful for bleeding disorder assessments should be translated into clinical practice.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.